University of Groningen
Definition of the sigma(W) Regulon of Bacillus subtilis in the Absence of StressZweers, Jessica C.; Nicolas, Pierre; Wiegert, Thomas; van Dijl, Jan; Denham, Emma L.
Published in:PLoS ONE
DOI:10.1371/journal.pone.0048471
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2012
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):Zweers, J. C., Nicolas, P., Wiegert, T., van Dijl, J. M., & Denham, E. L. (2012). Definition of the sigma(W)Regulon of Bacillus subtilis in the Absence of Stress. PLoS ONE, 7(11), [48471]. DOI:10.1371/journal.pone.0048471
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
Download date: 11-02-2018
Definition of the sW Regulon of Bacillus subtilis in theAbsence of StressJessica C. Zweers1, Pierre Nicolas2, Thomas Wiegert3, Jan Maarten van Dijl1*, Emma L. Denham1
1 Department of Medical Microbiology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands, 2 INRA, UR1077, Mathematique
Informatique et Genome, Jouy-en-Josas, France, 3 Hochschule Zittau/Gorlitz, FN/Biotechnologie, Zittau, Germany
Abstract
Bacteria employ extracytoplasmic function (ECF) sigma factors for their responses to environmental stresses. Despiteintensive research, the molecular dissection of ECF sigma factor regulons has remained a major challenge due to overlaps inthe ECF sigma factor-regulated genes and the stimuli that activate the different ECF sigma factors. Here we have employedtiling arrays to single out the ECF sW regulon of the Gram-positive bacterium Bacillus subtilis from the overlapping ECF sX,sY, and sM regulons. For this purpose, we profiled the transcriptome of a B. subtilis sigW mutant under non-stressconditions to select candidate genes that are strictly sW-regulated. Under these conditions, sW exhibits a basal level ofactivity. Subsequently, we verified the sW-dependency of candidate genes by comparing their transcript profiles totranscriptome data obtained with the parental B. subtilis strain 168 grown under 104 different conditions, including relevantstress conditions, such as salt shock. In addition, we investigated the transcriptomes of rasP or prsW mutant strains that lackthe proteases involved in the degradation of the sW anti-sigma factor RsiW and subsequent activation of the sW-regulon.Taken together, our studies identify 89 genes as being strictly sW-regulated, including several genes for non-coding RNAs.The effects of rasP or prsW mutations on the expression of sW-dependent genes were relatively mild, which implies that sW-dependent transcription under non-stress conditions is not strictly related to RasP and PrsW. Lastly, we show that thepleiotropic phenotype of rasP mutant cells, which have defects in competence development, protein secretion andmembrane protein production, is not mirrored in the transcript profile of these cells. This implies that RasP is not onlyimportant for transcriptional regulation via sW, but that this membrane protease also exerts other important post-transcriptional regulatory functions.
Citation: Zweers JC, Nicolas P, Wiegert T, van Dijl JM, Denham EL (2012) Definition of the sW Regulon of Bacillus subtilis in the Absence of Stress. PLoS ONE 7(11):e48471. doi:10.1371/journal.pone.0048471
Editor: Adam Driks, Loyola University Medical Center, United States of America
Received July 6, 2012; Accepted September 26, 2012; Published November 14, 2012
Copyright: � 2012 Zweers et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: JCZ, PN, JMvD, and ELD were in parts supported by the European Science Foundation under the EUROCORES Programme EuroSCOPE, grant 04-EScope01-011 from the Research Council for Earth and Life Sciences of the Netherlands Organization for Scientific Research, the CEU projects LSHM-CT-2006-019064,LSHG-CT-2006-037469, PITN-GA-2008-215524 and 244093, and by the transnational SysMO initiative through projects BACELL SysMO1 and 2. TW was supportedby the Deutsche Forschungsgemeinschaft (WI 1771/5-1). The funders had no role in study design, data collection and analysis, decision to publish, or preparationof the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: [email protected]
Introduction
Extracytoplasmic function (ECF) sigma factors enable bacteria
to respond adequately to harsh and stressful environmental
conditions. The numbers of ECF sigma factors vary among
different bacteria. While some bacteria (e.g. Mycoplasma genitalium)
have no ECF sigma factors, other bacteria can contain over 50
(Streptomyces coelicolor). In most cases however, only a limited
number of ECF sigma factors are present. For example, Escherichia
coli produces 2, and Bacillus subtilis 7 [23]. In non-stressed cells,
these sigma factors are usually inhibited by binding to a specific
anti-sigma factor [49]. For several anti-sigma factors it has been
shown that specific extracellular stresses trigger their regulated
intramembrane proteolysis (RIP) by site-1 and site-2 proteases in
the membrane [22,24,29,44]. Specifically, the site-1 protease clips
in the extracytoplasmic part of the anti-sigma factor and renders it
a substrate for the intramembrane cleaving site-2 protease. This
results in the release of the anti-sigma factor/sigma factor complex
into the cytoplasm, where the anti-sigma factor is further degraded
and the sigma factor can then redirect transcription
[10,21,23,27,29]. Attempts to accurately define each of the ECF
sigma factor regulons in organisms with multiple ECF sigma
factors have been complicated by partial overlaps that exist both
for the binding sites recognized by these sigma factors and the
stimuli that activate them. This is very clearly illustrated by studies
on the sW, sX, sY and sM sigma factors and their regulons in B.
subtilis [8,9,14,25,26,33,34,45]. To single out the individual ECF
sigma factor regulons is challenging, which is underscored by a
recent classification of the promoters of B. subtilis based on an
unsupervised algorithm [35]. This approach, which involved
transcript profiling across 104 different conditions, only allowed
the identification of a global ECF regulon, while the individual
sW, sX, sY and sM regulons remained undefined.
The sW regulon is among the three best-studied ECF sigma
factor regulons in B. subtilis. This regulon is induced in response to
cell envelope stress caused by antibiotics, alkaline shock and salt
shock [8,9,18,31,38,39,43,48]. The anti-sigma factor of sW, RsiW,
is cleaved by the site-1 protease PrsW and the site-2 protease RasP
[12,15,21,42,49]. Consistent with the requirement of PrsW for
RsiW degradation, prsW mutant cells have a phenotype that is
very similar to the phenotype of sigW mutant cells. In contrast,
PLOS ONE | www.plosone.org 1 November 2012 | Volume 7 | Issue 11 | e48471
deletion of the rasP gene causes a pleiotropic phenotype including
defects in the development of competence for genetic transforma-
tion and protein secretion [20,41]. Although transcriptional
analyses with sigW mutant cells were previously performed [8], a
detailed comparison of the effects of a sigW mutation with those of
prsW or rasP mutations on genome-wide transcription has not yet
been documented. Additionally, in the previous transcriptional
analyses of the sigW deletion strain, non-coding RNAs (ncRNAs)
were not included. Therefore, the present studies were aimed at
defining the strictly sW-regulated genes by transcript profiling
studies with tiling arrays using RNA from sigW, prsW or rasP
mutant strains. Notably, these array analyses were performed in
the absence of stress stimuli because, under these conditions sW
exhibits a basal well-detectable level of activity, while stress-related
side effects on the entire regulatory network are mostly absent.
The absence of stress thus provides a unique opportunity to obtain
an untroubled view of the sW regulon, even though sW-regulated
genes expressed at very low level might be missed. The results thus
obtained were enriched using data from the B. subtilis transcript
profiling study with tiling arrays in which gene expression in the
parental strain 168 was assessed under 104 different biological
conditions [35].
Materials and Methods
Bacterial strains, plasmids and growth conditionsThe bacterial strains and plasmids used in this study are listed in
Table 1. Strains were grown in Luria Bertani (LB) medium (Difco
Laboratories) at 37uC with vigorous shaking. Overnight grown
pre-cultures in LB medium were diluted to an OD600 of 0.05 in
fresh LB medium and then grown to the exponential phase as
determined by optical density readings. Under these conditions
sW is active but the cells are not stressed.
RNA isolationSamples for three biological replicates of each mutant and the
parental strain 168 were produced by independent culturing,
harvesting of the bacterial cells, and RNA isolation. When the
cultures reached an OD600 of 1.0 the equivalent of 15 OD units of
cells were harvested and total RNA was isolated according to
Eymann et al., 2002 [16] with some minor modifications. Cell
culture samples were added to 0.5 volume of frozen killing buffer
(20 mM Tris-HCl [pH 7.5], 5 mM MgCl2, 20 mM NaN3) and
centrifuged for 10 min at 4uC. The cell pellets thus obtained were
frozen in liquid nitrogen and stored at 280uC. Pellets were
resuspended in 200 ml ice-cold killing buffer and transferred to
precooled Teflon disruption vessels filled with liquid nitrogen.
Cells were then disrupted for 2 min at 2600 rpm in a Mikro-
Dismembrator S (Sartorius). The frozen powder was resuspended
in 4 mL prewarmed (50uC) lysis solution (4 M guanidine
thiocyanate, 25 mM sodium acetate [pH 5.2], 0.5% N-laurylsar-
cosinate [wt/vol]) and immediately frozen in liquid nitrogen. Total
RNA was isolated by acid-phenol extraction. Samples were
extracted twice with 1 volume of acid phenol/chloroform/isoamyl
alcohol (25:24:1, [pH 4.5]) and once with 1 volume of chloro-
form/isoamyl alcohol (24:1). After adding 1/10 volume of 3 M
sodium acetate (pH 5.2), RNA was precipitated overnight with
isopropanol at 220uC. Precipitated RNA was washed with 70%
ethanol and dissolved in 100 ml of RNase free water. The isolated
RNA was DNase-treated using the RNase-Free DNase Set
(Qiagen) and purified using the RNA Clean-Up and Concentra-
tion Micro Kit (Norgen). RNA concentrations were measured
using a Nanodrop-1000 spectrophotometer and RNA quality was
assessed with the Agilent 2100 Bioanalyzer according to the
manufacturer’s instructions. Labeling of the samples and hybrid-
izations were performed in strand-specific conditions by Nimble-
Gen, as previously described [40], using Basysbio_T2 tiling arrays
(NimbleGen). All tiling array data can be queried under the
NCBI-GEO accession numbers GSE35236 and GPL15150.
Statistical analysesAn aggregated expression measure was computed for each
annotated and for each transcribed segment recently identified in
the systematic study of transcriptome changes across lifestyles [35].
This measure consists of the median of the smoothed signal for
probes with a unique perfect match on the genome sequence lying
entirely within the boundaries of a particular feature [35]. The
data was quantile-normalized to remove trends caused by
technical variations between experiments [5]. A single linear
model was fitted on the log2-scale data to assess the links between
variations of expression and the genetic background of the
analyzed sigW, rasP or prsW mutant strains and the parental strain
168. The p-values associated with the tests for non-null effects of
each mutation compared to the parental strain were computed
(function ‘‘lm’’ in R). One of the three hybridizations for the prsW
mutant harbored an atypical transcriptome profile resembling that
of RNA extracted from stationary phase cells. We interpret this
observation as the result of a technical error when the samples
were prepared, and this data point was therefore discarded. From
the p-values, q-values allowing the control of the false discovery
rate were estimated using the procedure of Strimmer [3] as
implemented in the R package ‘‘fdrtool’’. To increase the
statistical power of our analyses, we also considered computation
of false discovery rates using the same procedure, but restricting
our attention to the subset of genes that were previously predicted
as part of the global ECF regulon [35].
Expression profiles across 104 conditions and ECF sigmafactor binding site predictions
In addition to our transcript profiling experiments with mutant
strains, we used the data from a study on the B. subtilis 168
transcriptome across 104 biological conditions (269 hybridiza-
tions), that was aimed at covering the maximum diversity of this
bacterium’s lifestyles [35]. These included growth on various
media and carbon-sources, responses to stresses and developmen-
tal processes, such as competence development and the sporula-
tion-germination cycle. In particular, we incorporated in our
analysis the newly identified transcription segments, such as
antisense RNAs and putative regulatory ncRNAs. For a high-level
comparison of expression profiles, we relied on a classification
based on average-linkage hierarchical clustering of the matrix of
pairwise correlation with a cut-off set to 0.4 that defined 167 high-
level clusters numbered in an arbitrary order C1 to C167. To
complement the list of genes previously reported as being
controlled by an ECF sigma factor, we also used the results of
an un-supervised classification of the sequences upstream tran-
Table 1. B. subtilis strains.
Strain Genotype Reference
168 trpC2 [35]
sigW trpC2 sigW::bleo, Bmr [42]
rasP trpC2 rasP::tc, Tetr [42]
prsW trpC2 prsW::bleo, Bmr [21]
doi:10.1371/journal.pone.0048471.t001
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 2 November 2012 | Volume 7 | Issue 11 | e48471
scription start sites that identified 79 putative ECF sigma-factor
dependent promoters [35].
Results
Two groups of down-regulated genes in sigW mutantcells
Several previously documented studies have employed different
strategies to identify genes that are regulated by sW
[2,6,8,9,25,26,34,46]. To accurately define the sW regulon and
to include possible ncRNAs that are controlled by sW under non-
stress conditions, we analyzed the genes that are down-regulated in
the sigW mutant compared to the parental strain with tiling arrays
(GEO accession number GSE35236). To ensure that genes not
related to sW activity were excluded from this study, we made use
of the fact that sW becomes active in the late exponential growth
phase under non-stress conditions [25]. This is important because
the absence of a stress stimulus provides a unique opportunity to
obtain an untroubled view of the sW regulon since stress-related
side effects on the entire regulatory network are largely absent. As
expected, most genes previously designated as part of the sW-
regulon were down-regulated in our tiling array analysis in the
sigW mutant compared to the parental strain. However, we
observed that the effect amplitudes varied considerably between
these genes, which allowed us to distinguish three subgroups
(Figure 1, Tables 2 and 3). Group 1 consists of genes that are
strongly down-regulated (this group has effect values ranging in
log2-scale from 24 to 21.5). The most strongly down-regulated
genes belonging to group 1 are rsiW and spo0M. Group 2 contains
previously reported sW-regulated genes that are less strongly
down-regulated due to the sigW mutation than the genes in group
1 (effect-values between 21.5 and 20.2). Group 3 consists of 16
genes that were previously reported as sW-regulated, but that
nonetheless were not down-regulated in the present transcriptome
analyses of the sigW mutant. Based on the present data, we
identified 89 potentially sW-regulated genes, which are located in
28 operons (Tables 2 and 3). The division of genes into groups 1
and 2 did not correlate with the transcription levels of these genes
in the parental strain (Mann-Withney U-test p-value = 0.23). This
rules out the possibility that the observed bimodal pattern of
down-regulation of genes in the sigW mutant is simply a reflection
of their transcription levels in the parental strain. Indeed, the
apparently bimodal down-regulation pattern of gene expression in
the sigW mutant probably results from more complex transcrip-
tional regulation. Of the 28 identified sW-regulated operons, 12
consist only of group 1 genes, and 8 consist only of group 2 genes.
In 8 operons a combination of group 1 and group 2 genes was
found, the group 2 genes always being localized at the end of these
operons. In many cases, the boundary between group 1 and 2
genes correlated with the presence of an internal promoter (before
yozO, ybfP, S161, yxjH, ydjO, S659, S716), or a terminator (after
ybfO, yvlD, ywrE, yqfB) that could potentially be responsible for
differences in their responses to the sigW deletion [35]. We also
examined the sequences corresponding to predicted ECF Sigma
factor binding sites [35] upstream of the genes of group 2 to those
of group 1, but could not identify differences in the sequences that
would explain the observed behavior.
Genes that were found to be statistically significantly down-
regulated in the sigW mutant are likely to be regulated by sW. To
establish this list of genes we computed q-values from the p-values,
which allowed us to control the number of false positive
identifications by taking into account the high number of genes
examined. Based on this statistical analysis, we propose that genes
down-regulated in the sigW mutant with q-values lower than 0.05
are most likely genuine sW-regulated genes (Table 2; genes with q-
values,0.05 are marked with *). However, if we consider only
these genes as being sW-regulated, several genes that were
previously shown to be sW-regulated by other methods (Table S1)
would have to be discarded from the sW regulon under non-stress
conditions despite their apparent down-regulation. To avoid such
Figure 1. Effect values for transcriptional changes in sigW mutant B. subtilis cells. The transcript abundance in sigW mutant cells wascompared to that in the parental strain 168 by tiling array analyses. The effect values were calculated on a log2 scale and the numbers of genes with aparticular effect value were plotted as a function of the effect values. The black line represents all analyzed genes. The dashed line represents only thegenes that are statistically significantly downregulated in the sigW mutant. The grey line represents the genes that were previously reported as beingsW-regulated. The groups 1, 2 and 3 of sW-regulated genes are indicated.doi:10.1371/journal.pone.0048471.g001
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 3 November 2012 | Volume 7 | Issue 11 | e48471
Table 2. Down-regulated genes in sigW mutant cells.
NameEffectsigW/WT Function Regulators
Geneticorganization Cluster
sWXY promotersequence Conclusion
rsiW 26.84* Control of sigW activity sW, AbrB sigW-rsiW C9 Yes core sW
sigW 26.83* Sigma W factor sW, AbrB sigW-rsiW C9 Yes core sW
spo0M 24.92* Sporulation sW, sH spo0M C9 Yes core sW
S691 23.66* S691-yoaG-S690 C9 Yes core sW
yeaA 23.61* sW, sE yeaA-ydjP-ydjO C2 Yes Secondary sW
ysdB 23.54* sW, sB ysdB C9 Yes core sW
yjoB 23.40* sW yjoB C9 Yes core sW
ydjP 23.34* sW, sE yeaA-ydjP-ydjO C2 Yes Secondary sW
S462 (indep) 23.23* C9 Yes core sW
yxjI 23.12* sW, sE, DegU S1495-yxjJ-yxjI C9 Yes core sW
yoaG 23.07* sW S691-yoaG-S690 C9 Yes core sW
fosB 23.03* Fosfomycin resistance sW fosB-S658-S659 C9 Yes core sW
ythP 22.98* ABC transporter(ATP binding protein)
sW ythP-ythQ C9 Yes core sW
S690 22.90* S691-yoaG-S690 C9 Yes core sW
S1495 (indep) 22.89* S1495-yxjJ-yjxI C9 Yes core sW
ythQ 22.74* ABC transporter sW ythP-ythQ C9 Yes core sW
S742 22.70* S742-yozO-S740-S739-yocM C9 Yes core sW
pspA 22.68* sW, AbrB pspA-ydjG-ydjH-ydjI C6 Yes Secondary sW
yfhL 22.52* SdpC resistance sW, sB yfhL-yfhM C5 Yes Secondary sW
ydjG 22.51* sW, AbrB pspA-ydjG-ydjH-ydjI C6 Yes Secondary sW
S719 (inter) 22.49* yobJ-S719 C9 Yes core sW
S658 (inter) 22.48* fosB-S658-S659 C9 Yes core sW
ybfO 22.47* sW, AbrB ybfO-ybfP-S89 C9 Yes core sW
ydbT 22.47* sW ydbS-ydbT-S160-S162-acpS C6 Yes Secondary sW
ydbS 22.46* sW ydbS-ydbT-S160-S162-acpS C6 Yes Secondary sW
pbpE 22.33* sW pbpE -racX C9 Yes core sW
yuaG (floT) 22.33* Sporulation(early stage)
sW yuaF-yuaG-yuaI C9 yes core sW
yfhM 22.30* Survival ofethanol stress
sW, sB yfhL-yfhM C5 Yes Secondary sW
ydjH 22.27* sW, AbrB pspA-ydjG-ydjH-ydjI C6 Yes Secondary sW
yqfB 22.25 Resistance againstsublancin
sW yqeZ-yqfA-yqfB-yqfC-yqfD C9 Yes core sW
yobJ 22.24* sW yobJ-S719 C9 Yes core sW
yqeZ 22.21 Serine protease,Resistance againstsublancin
sW yqeZ-yqfA-yqfB-yqfC-yqfD C9 Yes core sW
ydjI 22.17* sW, AbrB pspA-ydjG-ydjH-ydjI C6 Yes Secondary sW
racX 22.12* Control of biofilmformation
sW pbpE -racX C9 Yes core sW
yqfA 22.11 Resistance againstsublancin
sW yqeZ-yqfA-yqfB-yqfC-yqfD C9 Yes core sW
mtlF 22.05 Uptake of mannitol MtlR mtlA-mtlF-mtlD C36 No background
yuaI 22.02* sW yuaF-yuaG-yuaI C9 Yes core sW
mtlD 21.97 Mannitol utilization MtlR mtlA-mtlF-mtlD C36 No background
yvlA 21.91* sW, AbrB yvlA-yvlB-yvlC-yvlD-S1338 C9 Yes core sW
yvlB 21.85* sW, AbrB yvlA-yvlB-yvlC-yvlD-S1338 C9 Yes core sW
mtlA 21.85 Mannitol utilization MtlR mtlA-mtlF-mtlD C36 No background
ywrE 21.82 sW ywrE-S1390 C9 Yes core sW
yuaF 21.78* sW yuaF-yuaG-yuaI C9 Yes core sW
yoaF 21.58 sW yoaF C48 Yes Secondary sW
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 4 November 2012 | Volume 7 | Issue 11 | e48471
Table 2. Cont.
NameEffectsigW/WT Function Regulators
Geneticorganization Cluster
sWXY promotersequence Conclusion
S160 (inter) 21.56 ydbS-ydbT-S160-S162-acpS C6 Yes Secondary sW
ybfP 21.38* sW, AbrB ybfO-ybfP-S89 C9 Yes core sW
S89* 21.34* ybfO-ybfP-S89 C9 Yes core sW
yvlD 21.34 sW,AbrB yvlA-yvlB-yvlC-yvlD-S1338 C9 Yes core sW
yvlC 21.32 sW,AbrB yvlA-yvlB-yvlC-yvlD-S1338 C9 Yes core sW
yjzH 21.19 yjzH-S442 C9 Yes core sW
sppA 21.18 Signal peptidepeptidase
sW sppA-yteJ C6 Yes Secondary sW
yteJ 21.17* sW sppA-yteJ C6 Yes Secondary sW
yaaN 21.11 sW xpaC-yaaN-S22 C9 Yes core sW
yceE 21.04 Resistance againstethanol stress andcold
sW, sB, sM S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
S716 20.95 Downstream of yobJ-S719 C31 Yes Read through
S659 (indep) 20.94 fosB-S658-S659 C9 Yes core sW
yceD 20.90 Resistance againstethanol stress
sW, sB, sM S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
yceH 20.88* sW, sB, sM S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
S22 (intra) 20.88 xpaC-yaaN-S22 C9 Yes core sW
yceG 20.87* sW, sB, sM S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
yceC 20.84 sW, sB, sM S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
yxjH 20.83 S-box Downstream of S1495-yxjJ-yxjI C48 Yes Read through
ygzA 20.82 Opposite of spo0M C2 No Background
S1338 20.80 yvlA-yvlB-yvlC-yvlD-S1338 C9 Yes core sW
ilvD 20.78 Aminoacidbiosynthesis
CodY C48 No background
yknX 20.78 Resistance againstSdpC
sW,AbrB yknW-yknX-yknY-yknZ C9 Yes core sW
S106 20.78 S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
xpaC 20.77 sW xpaC-yaaN-S22 C9 Yes core sW
yqfC 20.76* sE yqeZ-yqfA-yqfB-yqfC-yqfD C2 Yes Secondary sW
yknY 20.76 Resistance againstSdpC
sW,AbrB yknW-yknX-yknY-yknZ C9 Yes core sW
S1175 20.75 59 mntA C1 No background
yceF 20.74 sW, sB, sM S106-yceC-yceD-yceE-yceF-yceG-yceH
C6 Yes Secondary sW
yqfD 20.72* sE yqeZ-yqfA-yqfB-yqfC-yqfD C2 Yes Secondary sW
yknZ 20.69 Resistance againstSdpC
sW,AbrB yknW-yknX-yknY-yknZ C9 Yes core sW
mtnK 20.65 S-box mtnK-mtnA C48 No background
alsD 20.63 alsS-alsD C39 No background
yozO 20.60 sW S742-yozO-S740-S739-yocM C9 Yes core sW
yknW 20.57 Resistance againstSdpC
sW,AbrB yknW-yknX-yknY-yknZ C9 Yes core sW
S740 (inter) 20.54 S742-yozO-S740-S739-yocM C6 Yes Secondary sW
S161 20.52 Fatty acid biosynthesis 59 acpS C3 Yes Secondary sW
S739 20.51 S742-yozO-S740-S739-yocM C2 Yes Secondary sW
S1390 (inter) 20.48 ywrE-S1390 C9 Yes core sW
S442 (inter) 20.48 yjzH-S442 C9 Yes core sW
acpS 20.45 Fatty acid biosynthesis ydbS-ydbT-S160-S162-acpS C3 Yes Secondary sW
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 5 November 2012 | Volume 7 | Issue 11 | e48471
potentially false negative exclusions, we maintained all the genes
that were down-regulated with q-values higher than 0.05 but p-
values lower than 0.05 also in our shortlist of potentially sW-
regulated genes. These genes were further analyzed by assessing
their transcription profiles under 104 conditions, including several
conditions known to induce SigW.
Definition of the sW regulon by assessment of transcriptprofiles across conditions
To minimize the false positive identifications of sW-regulated
genes, we took advantage of a large-scale tiling array analysis of
gene expression in B. subtilis 168 across 104 conditions, involving
269 hybridizations [35]; GEO accession number GPL15150).
Within this previous study promoters of different sigma factors
were classified based on an unsupervised algorithm. Notably, sW
regulated promoters were classified together with the other ECF
sigma factors (sW, sX, sY and sM) as having sigma factor binding
sites of the ‘sWXY’ type, because no distinction between promoters
recognized by sigma factors with similar DNA binding motifs
could be made (note that although this binding site was annotated
as ‘sWXY’ type, it also covers the sM binding site). Importantly,
the results of this study revealed marked differences in the
transcription profiles of the sigW, sigY, sigX and sigM genes across
conditions, especially during heat, salt and ethanol stress (Figure 2).
This was an important observation, because it can help in the
dissection of the respective regulons. The analysis of transcription
profiles across the 104 conditions showed that the transcription
profiles of 59 genes cluster with that of sigW in the previously
defined transcription cluster C9 (Figure 3, [35]). Importantly, most
genes in cluster C9 were found to be significantly down-regulated
in the sigW mutant in our present studies and/or were previously
reported as sW-regulated (Figure 4). The 12 genes within cluster
C9 that are not sW-regulated represent members of the sY
regulon, including the sigY gene itself. Their presence in cluster C9
relates to the fact that sY-regulated genes behave quite similarly to
sW-regulated genes, the main distinguishing feature being that
they are induced by ethanol stress rather than salt stress. Clearly,
the known sY-regulated genes in cluster C9 were not down-
regulated in the sigW mutant, whereas all other genes in cluster C9
were down-regulated in the sigW mutant (Figure 4A). Only one
gene in cluster C9, yxzE, which was previously reported to be sW-
regulated, did not qualify as a sW-regulated gene in our statistical
analyses as its down-regulation in the sigW mutant (effect value
20.45) had a p-value of 0.08. However, based on the combined
data, we believe that yxzE should be regarded as a member of the
sW regulon. Accordingly, the long 39 UTR of yxzE with the
designation S1489 is probably also part of the sW regulon, which
is supported by the fact that it is present in cluster C9 (Table 3).
20 genes that have previously been reported as sW-regulated
were also found to be down-regulated in the sigW mutant, but are
nevertheless not included in cluster C9 (Figure 4A). 14 of these
genes belong to cluster C6 (Figure 3), whereas the others are
distributed over several other clusters. Possibly, these genes are not
only regulated by sW, but also by other sigma factors or gene
regulators, which would lead to expression profiles that differ from
the sigW expression profile. Therefore, we examined the expres-
sion profiles of these genes with special attention to induction
during salt stress, which is a hallmark of the sW-regulated genes
[18,38,43]. In addition, we also compared these profiles with the
profiles of genes in the sM, sX and sY regulons that also respond
to cell envelope stress. These analyses revealed in total 79 genes
with ‘sW-like’ expression profiles that are induced upon salt stress
(i.e. 54 previously reported members of the sW regulon plus 25
newly identified sW-regulated genes; Figure 4B). Based on the
transcriptional profiles under different conditions, and the
requirement to be down-regulated in the sigW mutant, we propose
to make a distinction between core genes of the sW-regulon and
secondary sW-regulated genes. The sW-regulated genes in cluster
C9 would be the core genes of the sW regulon and all other sW-
dependent genes would be secondary sW regulon genes (Table 2).
The genes that were newly identified as being sW-regulated
were mainly novel ncRNAs that are part of sW-regulated operons
(Table 2). One distinct exception is the ncRNA S462, which is
located downstream of htrA. S462 is an independent ncRNA that is
preceded by a consensus sWXY binding site [35]. yjzH and the
downstream ncRNA S442 also represent novel members of the
sW regulon, which are preceded by a predicted sWXY binding
sequence. Additionally, in several occasions there was read-
through from sW-regulated operons into downstream genes. For
example, the operon yqeZ-yqfA-yqfB is known to be sW-regulated,
but the downstream genes yqfC and yqfD had previously not been
identified as being sW-regulated. Although yqfC and yqfD were not
as strongly down-regulated in the sigW mutant as the preceding
operon, the down-regulation of these genes was still clearly
significant with q-values of less than 0.05. Additionally, these genes
were found to be up-regulated during salt stress [35]. Therefore,
we conclude that yqfC and yqfD are truly sW-regulated. In other
cases of read through no induction during salt stress was observed,
and the respective genes are therefore not considered to be sW-
regulated.
Several genes further downstream of known sW-regulated
operons also behave like sW-regulated genes. Downstream of yozO
Table 2. Cont.
NameEffectsigW/WT Function Regulators
Geneticorganization Cluster
sWXY promotersequence Conclusion
S162 20.44 S162-ydcC C2 Yes Secondary sW
ydcC 20.42 sE S162-ydcC C2 Yes Secondary sW
thiC 20.41 Thiamine biosynthesis Thi-box Downstream of ygzA C48 No Background
ydjO 20.41 sW, sE yeaA-ydjP-ydjO C2 Yes Secondary sW
yocM 20.41 S742-yozO-S740-S739-yocM C2 Yes Secondary sW
Only the down-regulated genes with effect values lower than 20.4 and p-values lower than 0.05 are shown. Effect values marked with * have q-values of less than 0.05.For each individual gene, the Table lists the function, the previously identified regulation, the genetic organization, the condition-dependent transcription profile clusteras defined by Nicolas et al [35], the presence of a predicted ‘sWXY’ promoter sequence [35], and our conclusion whether it belongs to the sW core regulon or thesecondary sW-regulated genes. It should be noted here that the previously predicted ‘sWXY’ promoter sequence [35] also covers the potential binding site for sM. Thedivision between group 1 and group 2 genes is indicated by a bold line.doi:10.1371/journal.pone.0048471.t002
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 6 November 2012 | Volume 7 | Issue 11 | e48471
Ta
ble
3.
Pre
vio
usl
yR
ep
ort
eds
W-r
eg
ula
ted
ge
ne
sth
atw
ere
no
tsi
gn
ific
antl
yd
ow
n-r
eg
ula
ted
inth
esi
gW
mu
tan
tst
rain
.
Na
me
Eff
ect
sig
W/W
TF
un
ctio
nR
eg
ula
tio
nG
en
eti
co
rga
niz
ati
on
Clu
ste
rs
WX
Yp
rom
ote
rse
qu
en
ceC
on
clu
sio
n
yxzE
20
.45
sW
,Ab
rByx
zE-S
14
89
C9
Ye
sco
res
W
S14
89
20
.31
yxzE
-S1
48
9C
9Y
es
core
sW
bsc
R(f
atR
)2
0.1
5Fa
tty
acid
bio
syn
the
sis
sM
,s
W,s
X,
FatR
yrh
H-b
scR
-yrh
JC
10
Ye
ss
M
ywb
O2
0.0
2Ir
on
up
take
sM
,s
W,s
X,
Fur
ywb
N-y
wb
OC
29
No
No
ts
W-
Fur-
reg
ula
ted
fab
Ha
20
.03
Fatt
yac
idb
iosy
nth
esi
ss
W,
Fap
Rfa
bH
A-f
ab
FC
3Y
es
(P5
+11
4)
sW
Kin
gst
on
et
al2
01
1
efeN
(yw
bN
)2
0.0
4s
M,s
W,s
X,
Fur
ywb
L-yw
bM
-yw
bN
ywb
N-y
wb
OC
29
No
No
ts
W-
Fur-
reg
ula
ted
fab
F2
0.0
7Fa
tty
acid
bio
syn
the
sis
sW
,Fa
pR
fab
HA
-fa
bF
C3
Ye
ss
WK
ing
sto
ne
tal
20
11
yrh
J(c
ypB
)2
0.0
8s
M,s
W,s
Xyr
hH
-fa
tR-y
rhJ
C1
0Y
es
sM
ywa
C0
.06
(p)p
pG
pp
syn
the
tase
sM
,s
WC
79
Ye
ss
M
yjb
C0
.07
Pe
rR,s
B,s
M,s
W,s
Xyj
bC
-sp
xC
5Y
es
sB
ands
M
yjb
D(s
pxA
)0
.17
Pe
rR,s
B,s
M,s
W,s
Xyj
bC
-sp
xC
17
Ye
ss
Ban
ds
M
div
IC0
.23
Sep
tum
form
atio
ns
E,s
M,s
W,s
Xya
bM
-ya
bN
-ya
bO
-ya
bP
-yq
ab
Q-
div
IC-y
ab
RC
20
Ye
ss
M
yrh
H0
.25
sM
,s
W,s
Xyr
hH
-fa
tR-y
rhJ
C1
0Y
es
sM
ab
h0
.30
Ge
ne
reg
ula
tio
nd
uri
ng
tran
siti
on
ph
ase
sW
,s
XC
36
Ye
ss
X
ywn
J0
.31
sF,s
M,s
W,s
XC
2Y
es
sM
bcr
C0
.33
Re
sist
ance
tob
acit
raci
nan
do
xid
ativ
est
ress
sI ,s
M,s
W,s
XC
10
Ye
ss
M
yqjL
0.3
7R
esi
stan
ceag
ain
stp
araq
uat
sB,s
M,s
WC
78
Ye
ss
B,s
M
For
eac
hin
div
idu
alg
en
e,t
he
Tab
lelis
tsth
efu
nct
ion
,th
ep
revi
ou
sly
ide
nti
fie
dre
gu
lati
on
,th
eg
en
eti
co
rgan
izat
ion
,th
eco
nd
itio
n-d
ep
en
de
nt
tran
scri
pti
on
pro
file
clu
ste
ras
de
fin
ed
by
Nic
ola
set
al
[35
],th
ep
rese
nce
of
ap
red
icte
d‘s
WX
Y’p
rom
ote
rse
qu
en
ce[3
5],
and
ou
rco
ncl
usi
on
wh
eth
er
itb
elo
ng
sto
thes
Wco
rere
gu
lon
or
the
seco
nd
arys
W-r
eg
ula
ted
ge
ne
s.It
sho
uld
be
no
ted
he
reth
atth
ep
revi
ou
sly
pre
dic
ted
‘sW
XY’p
rom
ote
rse
qu
en
ce[3
5]
also
cove
rsth
ep
ote
nti
alb
ind
ing
site
fors
M.
do
i:10
.13
71
/jo
urn
al.p
on
e.0
04
84
71
.t0
03
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 7 November 2012 | Volume 7 | Issue 11 | e48471
for example, S740, S739 and yocM are all down-regulated in the
sigW mutant and induced upon salt stress (Figure 5A). Similarly,
downstream of the ydbST operon, S161, acpS and S162 are down-
regulated in the sigW mutant and induced upon salt stress
(Figure 5B). In other cases the situation is different. For example,
ygzA, a gene starting close to the start site of spo0M, but running in
the opposite direction, is also down-regulated in the sigW mutant.
Nevertheless, ygzA is not preceded by a consensus binding
sequence for sWXY, and this gene is also not induced by salt
stress. Likewise, the yxjH gene downstream of the sW-regulated
gene yxjI is down-regulated in the sigW mutant, but also in this case
no induction is observed during salt stress. Thus, we do not
consider ygzA and yxjH to be genuinely sW-regulated genes.
15 genes that were previously reported to be sW-regulated were
not down-regulated in the sigW mutant (Table 3, Figures 1, 3 and
4). This observation cannot be explained by a simple absence of
expression of these genes in the parental strain that would have
precluded the possibility to observe their down-regulation. This
view is supported by the finding that the distribution of the
expression levels of these genes in the parental strain was not
significantly different from the distribution of the expression levels
of genes belonging to groups 1 and 2 (Mann-Withney U-test p-
value of 0.64). Indeed, these genes have been assigned to multiple
s regulons besides the sW regulon and they mostly appear to show
condition-dependent transcription profiles that are more similar to
those of genes regulated by s factors other than sW (Table 3). We
therefore examined whether these genes had been previously
shown to act in a typical sW-dependent manner, or whether their
dependency on other ECF sigma factors had been shown (Table
S1). The majority of these 15 genes do not show a typical
upregulation pattern under conditions inducing the sW regulon.
11 of the 15 genes have been shown to be regulated by other sigma
factors (10 by sM and 1 by sX). ywnJ, ywbN and yrhH have only
been shown to have the potential for binding sW in vitro [7,8,25],
and no in vivo data suggest a sW-dependence of their promoters.
fabHa has been shown to be expressed sW-dependently [31], and
upregulation of the fabHa-fabF operon has been reported upon
overexpression of sW [2]. However, this operon was never
observed to be upregulated in any of the conditions known to
induce the sW regulon. This is somewhat surprising, but may be
explained by the promoter being located within the fabHa gene
itself. The majority of these 15 genes are therefore unlikely to be
sW-regulated.
Lastly, 40 genes appeared to be up-regulated in the sigW mutant
with effect values of more than 0.4 and p-values of less than 0.05
(Table 4). However, it should be noted that none of these changes
have q-values smaller than 0.05. This suggests that these up-
regulations may represent false positive results or indirect effects
that are not as strong as direct regulatory effects. Several of the up-
regulated genes are located in the close proximity of sW-regulated
genes, but are encoded by the opposite strand. Two of these genes,
ybbK and ybbJ, are located immediately opposite of sigW and,
Figure 2. Expression profiles of sigW, sigX, sigY and sigM in B. subtilis 168 across 104 conditions. The 269 tiling array hybridizations [35] arearranged along the x-axis. Of particular interest for discriminating the activities of the encoded sigma factors are the conditions heat stress (‘heat’),ethanol stress (‘etha’) and hypersaline stress (‘salt’), which are marked by pink shading.doi:10.1371/journal.pone.0048471.g002
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 8 November 2012 | Volume 7 | Issue 11 | e48471
therefore, the up-regulation of these genes in the sigW mutant
could be the result of a polar effect of the deletion of sigW.
However, the transcription profiles of both of these genes do not
show changes during exposure to high salt and the same is true for
the other up-regulated genes. Therefore, we do not consider ybbK,
ybbJ and other genes up-regulated in the sigW mutant as novel sW-
regulated genes.
Function of the sW regulonThe sW regulon is responsible for activating genes whose
products are likely to be needed upon envelope stress, or beneficial
under conditions of alkali shock, salt stress and treatment with
cationic peptides and agents that impair cell wall biosynthesis
[9,18,38,39,43]. To verify this view, the genes identified in our
study as being part of the sW regulon were analysed for function
according to their classification in SubtiWiki [17] (Table 2, Table
S1). Indeed, the groups of genes that were most largely represented
encoded cell envelope stress proteins, membrane proteins and
proteins involved in resistance against toxins or antibiotics. These
proteins have been implicated in protecting the cell from stresses
that affect the membrane and in detoxification upon contact with
toxic compounds. Our present findings suggest that, also under
non-stress conditions, it may be beneficial for B. subtilis to express
the respective sW-regulated genes at a basal level, for example to
allow fast and effective responses to any membrane stresses that
may suddenly occur. Notably, over half of the genes identified as
being sW-regulated are B. subtilis ‘y’ genes, essentially genes that
have yet to be functionally annotated. Therefore, until the
functions of these genes are defined it will remain difficult to
determine which sW-regulated genes function in what capacity
when the regulon is upregulated.
Comparison of global transcription in rasP, prsW andsigW mutant cells
Deletion of the genes for RasP and PrsW under stress conditions
inhibits the activation of the sW-regulon, because both of these
proteases are required for inactivation of the sW anti-sigma factor
RsiW. Thus, no activation of sW-controlled genes was detectable
in rasP or prsW mutant cells upon stress [15,19,21,42]. In addition,
the rasP mutant is known to display several phenotypes, such as
defects in competence and protein secretion, which are not
observed in prsW or sigW mutants [20,32,41,47]. During
membrane protein overproduction, the rasP mutant also behaves
differently from the prsW and sigW mutants. Whereas prsW and
sigW mutations generally improve membrane protein overpro-
duction, in the rasP mutant overproduction of all tested membrane
proteins was abolished [50].
We wanted to know whether RasP and PrsW, the genes of
which are both expressed under the tested non-stress conditions,
play a role in the control of the basal activity of the sW regulon.
Generally, the transcriptional changes in the rasP or prsW mutant
strains compared to the parental strain and the sigW mutant were
rather small and only few had q-values below 0.05 (15 in the rasP/
WT comparison, 0 in the prsW/WT comparison, 21 in the rasP/
sigW comparison, and 14 in the prsW/sigW comparison). Closer
examination revealed that only 3 genes associated with q-values
below 0.05 were not predicted to belong to the global ECF regulon
defined in Nicolas et al. [35] (i.e. natA, hisG and tetB). We therefore
reasoned that statistical power could be increased by searching for
Figure 3. Assignment of clusters of genes with related transcript profiles across conditions to different groups of genes that aredown-regulated or up-regulated in sigW mutant cells. The down-regulated genes are represented by groups 1 and 2 (see also Fig. 1). Genes ingroup 3 were previously reported as sW-regulated, but our present studies provided no evidence for their proposed sW-dependency (see Fig. 1). Theup-regulated genes are represented in a separate bar. Previously defined transcription clusters [35] are indicated in each bar by their C-number.doi:10.1371/journal.pone.0048471.g003
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 9 November 2012 | Volume 7 | Issue 11 | e48471
differential expression in priority among the 243 genes and new
expression segments included in this analysis that were previously
classified as members of the global ECF regulon [35]. Indeed, the
estimates that we obtained for the false discovery rates of global
ECF regulon genes with p-values#0.05 were 9.7% for the sigW/
WT comparison, 10.3% for the rasP/WT comparison, 12.3% for
the prsW/WT comparison, 11.8% for the rasP/sigW comparison
and 13.4% for the prsW/sigW comparison. These genes are listed
in Tables S2, S3, S4. For completeness, other genes with p-
values#0.05 have also been listed although they probably include
a much higher fraction of false discoveries. Altogether, the
composition of these lists revealed that the afore-described sW-
regulated genes were down-regulated in both the rasP and prsW
mutants, but to lesser extents than in the sigW mutant. This
indicates that the deletion of rasP or prsW indeed decreased the
activity of sW, but that sW activity was not completely abolished
in the respective mutants under the applied non-stress conditions
(Figure 6, Table S2A). Apparently, some sW molecules were able
to escape from binding to RsiW, even in the absence of RasP or
PrsW, thereby causing low-level expression of the sW regulon that
was independent from intramembrane proteolysis by RasP and
PrsW. Among the non-sW-regulated genes that were down-
regulated in the rasP mutant were several genes that are involved
in the development of genetic competence (i.e. oppA, nucA, ssbB,
rapD). Other genes that were specifically down-regulated in the
rasP mutant mainly relate to lipid and cell wall turnover.
In both the rasP and prsW mutant strains, slight increases in
transcription were detected for genes involved in compatible solute
transport, which is important for osmoregulation (Table S2). Even
though not all of these genes were always significantly up-regulated
in each mutant, there seemed to be a mild, general up-regulation
of these genes in both the rasP and prsW mutant strains.
Additionally, slightly increased transcription of genes involved in
teichoic acid synthesis, phospholipid biosynthesis, cell wall
biogenesis and cell shape was observed. Genes that were
specifically up-regulated in the rasP mutant include genes involved
in amino acid metabolism (e.g. genes for histidine and arginine
biosynthesis, and ornithin and citrullin utilization) and genes
involved in cell envelope stress systems (e.g. the natAB-yccK operon
[11,36,37], the LiaRS, WalRK [4,13] and DesRK two-component
systems, and the sM-regulon [14,28,34]). However, not all genes
regulated by these systems were up-regulated and therefore the
significance of these findings remains unclear.
Notably, in our previous studies we have reported significantly
increased levels of HtrA and HtrB in the rasP mutant [50].
Nevertheless, the cssR and cssS transcription levels were only
slightly down-regulated in the rasP mutant and the same was true
for the sigW or prsW mutant strains (Table 5). Furthermore, the
transcription of the CssRS-regulated htrA and htrB genes was not
significantly altered in rasP, prsW or sigW mutant cells (Table 5).
This implies that the activity of the CssRS system is not responsible
for the increased HtrA and HtrB levels in the rasP mutant.
Lastly, a direct comparison of global transcription in the rasP
and sigW mutant strains resulted in very few statistically significant
changes (Tables S3 and S4). Compared to the sigW mutant, a few
Figure 4. Venn diagrams for the comparison of genes that werefound to be downregulated in the sigW mutant strain withpreviously reported sW-regulated genes and genes thatdisplay similar condition-dependent transcription profiles assigW. Diagram A includes only the so-called cluster C9 genes that havehighly similar condition-dependent transcription profiles as defined byNicolas et al [35]. Notably, the sigW gene is included in cluster C9.Diagram B includes all genes that show condition-dependentexpression profiles similar to that of sigW, including induction uponsalt stress.doi:10.1371/journal.pone.0048471.g004
Figure 5. Organization of complex sW-regulated operons. The sW-regulated ORFs are indicated in black, and the sW-regulated ncRNAs areindicated in grey. Genes and an ncRNA on the opposite strand are indicated in white. A, The yozO-yocM operon. B, The ydbST-acpS operon.doi:10.1371/journal.pone.0048471.g005
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 10 November 2012 | Volume 7 | Issue 11 | e48471
Table 4. Genes that were up-regulated in the sigW mutant.
Name Effect sigW/WT Function Regulation Genetic organization Cluster
ybbK 3.07 ybbK-ybbJ Opposite of sigW (Q) C6
ybbJ 2.68 ybbK-ybbJ Opposite of sigW (Q) C6
S928 (inter) 2.25 Between mgsR and rsbRD C5
S1380 1.16 C10
ykzV 1.13 C2
S1026 (inter) 0.92 Upstream of yrzI (q) C2
cotT 0.91 C2
yodI 0.83 sK C2
S1030 0.82 59 of yrhF, C3
murG 0.77 Peptidoglycan precursorbiosynthesis
sE, sM,SpoIID C4
S981 0.73 39 of yqaP, opposite ofyqaR (q) and yqbC (q)
C17
ymaG 0.68 Spore coat protein sK C2
S655 0.66 Opposite of fosB (Q) C17
S862 0.65 59 of spoIVA C2
S1356 0.64 59 of degS (q) C3
yrzI 0.63 C2
S613 0.62 59 of ymzD (slightly q) C27
S663 0.61 59 of ccdA (slightly q) C17
S1405 (inter) 0.60 Downstream ofspoIID (slightly q)
C2
S254 (indep) 0.60 C17
ykzW 0.59 RNA that inhibits AhrC translation CcpN regulon C30
S653 (indep) 0.57 Downstream of fosB (Q) C17
ydeH 0.56 AbrB C17
yqaR 0.54 Close to S981 (q)and yqbC (q) C6
S360 (inter) 0.54 C35
S118 (inter) 0.52 Opposite of yuaI-yuaF-yuaG C52
obg 0.50 Ribosome assembly (essential),possibly required forSpo0A-activation
C3
cotU 0.50 Spore coat protein GerE, GerR C2
yqxD 0.46 sH upstream of S952 (slightly q) C153
S278 0.46 59 yfzA (q) C17
pssA 0.46 Biosynthesis of phospholipids Upstream of ybfO-ybfP(Q) C3
S303 0.45 59 of ygxA C3
comK 0.44 Competence and DNA uptakeregulation
AbrB, ComK,DegU, CodY, Rok
C1
yktD 0.43 C115
S1543 (intra) 0.43 Upstream of yydI, yydJ(both slightly q)
C35
S95 0.42 59 of ycbJ C35
S831 0.42 59 of ypeP C2
S427 0.42 59 of yjzE C2
S924 0.41 59 of sinI C17
yfzA 0.41 S278(q)-yfzA C17
Only the genes with Effect values higher than 0.4 and p-values lower than 0.05 are shown. Arrows behind genes in the ‘genetic organization’ column indicate whetherthe transcription of these genes was up- (q) or down-regulated (Q). For each individual gene, the Table lists the function, the previously identified regulation, thegenetic organization, and the condition-dependent transcription profile cluster as defined by Nicolas et al [35].doi:10.1371/journal.pone.0048471.t004
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 11 November 2012 | Volume 7 | Issue 11 | e48471
genes including rocD and rocA, natA and natB, des and argI were
specifically up-regulated in the rasP mutant. Other transcriptional
changes summarized in Table S3A relate to changes in the sigW
mutant. For the genes that were down-regulated in the rasP
mutant, most hits were specific for the rasP mutant. No clear
pattern however emerges from these changes, although some of
these genes relate to the cell envelope metabolism (membrane and
cell wall). Furthermore, the vast majority of genes found to be
differentially expressed in the prsW mutant compared to the sigW
mutant relate to sW-regulated genes. Only the up-regulation of
the pstS, pstBA, pstBB, pstA and pstC genes for phosphate uptake was
very specific for the prsW mutant. The reasons for these specific
differences in transcription in the rasP, prsW or sigW mutant strains
remains to be determined.
Discussion
The sW-regulon has been extensively described in several
previous papers, and 69 genes have been reported as sW-
controlled genes [8,9,25,26,31]. However, it has so far remained
very difficult to discriminate between genes of the sW-regulon and
the other ECF s-regulons of B. subtilis, as the respective promoter
sequences and the stress stimuli for induction partially overlap
[9,14,25]. Indeed, in the study reporting the transcriptional profile
of B. subtilis grown in 104 conditions [35], only a global ECF sigma
factor regulon was described, and no clear definition of the sW
regulon could be generated. Also, it was so far unknown which
ncRNAs of B. subtilis are part of the sW-regulon. In our present
studies, we have therefore employed tiling array data to define the
transcriptome of a sigW mutant B. subtilis strain. Then the results
were examined in the light of the recently described transcriptome
of the parental strain 168 across 104 different conditions [35]. Our
results show that 89 genes of B. subtilis are regulated by sW and the
data suggest that 13–15 of the 69 previously reported sW-
regulated genes might represent false-positive identifications. In
addition to 53 already known sW-regulated genes, we have
discovered 36 novel genes of the sW-regulon and we found that
several sW-regulated operons are larger than initially thought.
Two subgroups of sW-regulated genes can be discerned based
on the effect values for their down-regulation in sigW mutant cells.
This differential down-regulation pattern does not correlate with
the expression levels of these genes in the parental strain.
However, there appears to be a bias for genes that are located
at the downstream ends of certain large operons that often have
low effect values (i.e. group 2 genes), whereas the genes located
more upstream in these operons tend to have high effect values
(group 1 genes). On the other hand, several complete operons
display high effect values from start to end, while other complete
operons have low effect values from start to end. This indicates
that the location of a gene in an operon can influence whether it
belongs to group 1 or group 2. However, it remains to be
determined which additional mechanisms are responsible for the
observed bimodal pattern in sW regulation. Another novel finding
was that several apparently non-sW-regulated genes on the
opposite strand of sW-regulated genes turned out to be slightly
up-regulated in the sigW mutant. This indicates that the
transcriptional activity of sW-regulated genes can have a negative
impact on the transcription of genes encoded by the opposite
strand. The molecular basis for this effect is currently not known.
However, it is conceivable that RNA-polymerase initiating with
sW may directly or indirectly dampen the transcription elongation
efficiency of RNA-polymerase transcribing into the opposite
direction.
As expected, the sW-regulated genes were also down-regulated
in rasP or prsW mutant strains, albeit to lesser extents than in the
sigW mutant. This implies that there is residual sW activity in the
absence of either the RasP or PrsW proteases, which may relate to
the equilibrium between the free states of sW plus RsiW and the
sW-RsiW bound state. Such leakiness is not an uncommon feature
among biological systems. Alternatively, certain other proteases
may also be capable of degrading limited amounts of RsiW in the
absence of RasP or PrsW. Candidate proteases for alternative
Figure 6. Up- and down-regulation of genes in rasP, prsW orsigW mutant strains compared to the wild-type. A, Venn diagramfor down-regulated genes. B, Venn diagram for upregulated genes.Only genes with transcriptional changes that have p-values lower than0.05 and effect values lower than 20.40 (A) or higher than 0.40 (B) areincluded. The genes that are considered to be sW-regulated areindicated between brackets.doi:10.1371/journal.pone.0048471.g006
Table 5. Transcriptional changes of genes regulated by theCssRS two-component system.
Effect sigW/WT Effect rasP/WT Effect prsW/WT
cssR 20.20 20.20 20.19
cssS 20.22 20.22 20.03
htrA 20.24 0.21 0.07
htrB 20.29 0.18 0.06
doi:10.1371/journal.pone.0048471.t005
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 12 November 2012 | Volume 7 | Issue 11 | e48471
RsiW cleavage in the absence of PrsW might be the membrane-
bound forms of HtrA and HtrB. Both HtrA and HtrB are closely
related to the site-1 protease DegS of E. coli, which has been
implicated in RIP of the anti-sigma factor RseA that sequesters sE
[1,12,30]. It should be noted that, compared to the previously used
methods for assessing the effects of mutations in rasP or prsW
[15,19,21,42], the presently performed tiling array analyses are
more sensitive and they can reproducibly reveal smaller changes.
This is probably the reason why residual sW activity in the
absence of RasP or PrsW has so far remained unnoticed.
In relation to the previously documented defects of rasP mutant
cells in competence development [20,32], protein secretion
[32,41], and membrane protein overproduction [50], we verified
whether any of these defects could be connected to transcriptional
changes. However, as indicated above, the observed transcrip-
tional changes in the rasP mutant were generally very minor and,
apart from four competence-related genes, no changes were found
that might explain any of the observed phenotypes through
transcriptional regulation. The four affected competence-related
genes (nucA, oppA, ssbB and rapD) were only very slightly down-
regulated in the rasP mutant and this finding should be viewed
with caution, because the present analyses were performed with
cells grown in LB medium, which is not an optimal medium for
inducing competence. Taken together, we conclude that the
observed defects of rasP mutant cells in protein secretion and
membrane protein overproduction most likely relate to post-
transcriptional regulatory mechanisms that would involve the
enzymatic activity of the RasP protease. However, we cannot
completely exclude the possibility that changes in the membrane
fluidity contribute to the pleiotropic phenotype of rasP mutant
cells. This relates to recent studies by Kingston et al., 2011 [31],
who proposed that activation of a sW-dependent promoter in the
fabHa-fabF operon results in a higher proportion of straight-chain
fatty acids and a longer average chain length in phospholipids,
which will cause a reduced fluidity of the membrane. It should be
noted however that under non-stress conditions we observed no
influence of the absence of sW on the expression of fabHa.
In conclusion, the present studies lead to a definition of the sW
regulon under non-stress conditions (exponential growth in LB
broth at 37uC) that have been applied in numerous studies over
the past decades. Importantly, the use of non-stress conditions
allowed us to determine the basal expression levels of sW-
regulated genes, and to avoid side effects of particular stresses on
the entire regulatory network of the cell. By following this strategy,
we have considerably reduced the complexity of the system, which
permitted us (i) to pinpoint the most strictly sW-dependent genes
that probably have promoter sequences with the highest affinity
for sW, and (ii) to classify the known and newly identified sW-
controlled genes. Furthermore, our studies provide novel insights
in the importance of the RIP proteases PrsW and RasP in the
activation of this stress-responsive regulon. Especially, the obser-
vation that the absence of either PrsW or RasP does not lead to a
complete inactivation of sW-dependent gene expression is
intriguing and calls for further investigations. Although this
expression is most likely caused by an equilibrium where low
levels of sW bind to RNAP instead of the anti-sigma factor RsiW,
it cannot be excluded that certain, so far unknown, signals trigger
alternative pathways for RsiW inactivation, or that PrsW and
RasP might be substituted to some extent by other proteases.
Lastly, our present findings strongly support the view that RasP is
not only directly involved in the activation of the sW-regulon, but
also in other post-transcriptional regulatory mechanisms relating
to competence development, protein secretion and membrane
protein biogenesis.
Supporting Information
Table S1 Previously identified sW-regulated genes.
(XLSX)
Table S2 Genes down- or up-regulated in sigW, prsW or rasP
mutant strains. Changes associated with p-values,0.05 are
indicated in bold. A, down-regulated genes. B, up-regulated genes.
(DOCX)
Table S3 Genes up- or down-regulated in the rasP mutant strain
compared to the sigW mutant strain. Changes associated with p-
values,0.05 are indicated in bold. A, up-regulated genes. B,
down-regulated genes.
(DOCX)
Table S4 Genes up- or down-regulated in the prsW mutant
strain compared to the sigW mutant strain. Changes associated
with p-values,0.05 are indicated in bold. A, up-regulated genes.
B, down-regulated genes.
(DOCX)
Acknowledgments
The authors thank Ulrike Mader for helpful discussions and support.
Author Contributions
Conceived and designed the experiments: JCZ JMD ELD. Performed the
experiments: JCZ ELD. Analyzed the data: JCZ PN ELD. Contributed
reagents/materials/analysis tools: TW JMD. Wrote the paper: JCZ PN
TW JMD ELD.
References
1. Alba BM, Leeds JA, Onufryk C, Lu CZ, Gross CA (2002) DegS and YaeL
participate sequentially in the cleavage of RseA to activate the sigma(E)-
dependent extracytoplasmic stress response. Genes Dev 16: 2156–2168.
2. Asai K, Yamaguchi H, Kang CM, Yoshida K, Fujita Y, et al (2003) DNA
microarray analysis of Bacillus subtilis sigma factors of extracytoplasmic function
family. FEMS Microbiol Lett 220: 155–160.
3. Strimmer K (2008) A unified approach to false discovery rate estimation. BMC
Bioinformatics 9: 303.
4. Bisicchia P, Lioliou E, Noone D, Salzberg LI, Botella E, et al (2010),
Peptidoglycan metabolism is controlled by the WalRK (YycFG) and PhoPR two-
component systems in phosphate limited Bacillus subtilis cells. Mol Microbiol 75:
972–989.
5. Bolstad BM, Irizarry RA, Astrand M, Speed TP (2003) A comparison of
normalization methods for high density oligonucleotide array data based on
variance and bias. Bioinformatics 19: 185–193.
6. Butcher BG, Helmann JD (2006) Identification of Bacillus subtilis sigma-
dependent genes that provide intrinsic resistance to antimicrobial compounds
produced by Bacilli. Mol Microbiol 60: 765–782.
7. Cao M, Helmann JD (2004) The Bacillus subtilis extracytoplasmic-function
sigmaX factor regulates modification of the cell envelope and resistance to
cationic antimicrobial peptides. J Bacteriol 186: 1136–1146.
8. Cao M, Kobel PA, Morshedi MM, Wu MF, Paddon C, et al (2002) Defining the
Bacillus subtilis sigma(W) regulon: a comparative analysis of promoter consensus
search, run-off transcription/macroarray analysis (ROMA), and transcriptional
profiling approaches. J Mol Biol 316: 443–457.
9. Cao M, Wang T, Ye R, Helmann JD (2002), Antibiotics that inhibit cell wall
biosynthesis induce expression of the Bacillus subtilis sigma(W) and sigma(M)
regulons. Mol Microbiol 45: 1267–1276.
10. Chen G, Zhang X (2010) New insights into S2P signaling cascades: regulation,
variation, and conservation. Protein Sci 19: 2015–2030.
11. Cheng J, Guffanti AA, Krulwich TA (1997) A two-gene ABC-type transport
system that extrudes Na+ in Bacillus subtilis is induced by ethanol or
protonophore. Mol Microbiol 23: 1107–1120.
12. Dalbey RE, Wang P, van Dijl JM (2012) Membrane proteases in the bacterial
protein secretion and quality control pathway. Microbiol Mol Biol Rev 76: 311–
330.
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 13 November 2012 | Volume 7 | Issue 11 | e48471
13. Dubrac S, Bisicchia P, Devine KM, Msadek T (2008) A matter of life and death:
cell wall homeostasis and the WalKR (YycGF) essential signal transductionpathway. Mol Microbiol 70: 1307–1322.
14. Eiamphungporn W, Helmann JD (2008) The Bacillus subtilis sigma(M) regulon
and its contribution to cell envelope stress responses. Mol Microbiol 67: 830–848.
15. Ellermeier CD, Losick R (2006) Evidence for a novel protease governingregulated intramembrane proteolysis and resistance to antimicrobial peptides in
Bacillus subtilis. Genes Dev 20: 1911–1922.
16. Eymann C, Homuth G, Scharf C, Hecker M (2002) Bacillus subtilis functionalgenomics: global characterization of the stringent response by proteome and
transcriptome analysis. J Bacteriol, 184:2500–2520.17. Florez LA, Roppel SF, Schmeisky AG, Lammers CR, Stulke J (2009) A
community-curated consensual annotation that is continuously updated: theBacillus subtilis centred wiki SubtiWiki, Database (Oxford): bap012.
18. Hahne H, Mader U, Otto A, Bonn F, Steil L, et al (2010) A comprehensive
proteomics and transcriptomics analysis of Bacillus subtilis salt stress adaptation.J Bacteriol 192: 870–882.
19. Heinrich J, Hein K, Wiegert T (2009) Two proteolytic modules are involved inregulated intramembrane proteolysis of Bacillus subtilis RsiW. Mol Microbiol
74: 1412–1426.
20. Heinrich J, Lunden T, Kontinen VP, Wiegert T (2008) The Bacillus subtilisABC transporter EcsAB influences intramembrane proteolysis through RasP.
Microbiology 154: 1989–1997.21. Heinrich J, Wiegert T (2006) YpdC determines site-1 degradation in regulated
intramembrane proteolysis of the RsiW anti-sigma factor of Bacillus subtilis. MolMicrobiol 62: 566–579.
22. Heinrich J, Wiegert T (2009) Regulated intramembrane proteolysis in the
control of extracytoplasmic function sigma factors. Res Microbiol 160: 696–703.23. Helmann JD (2002) The extracytoplasmic function (ECF) sigma factors. Adv
Microb Physiol 46: 47–110.24. Ho TD, Ellermeier CD (2012) Extra cytoplasmic function sigma factor
activation. Curr Opin Microbiol 15: 182–188.
25. Huang X, Fredrick KL, Helmann JD (1998) Promoter recognition by Bacillussubtilis sigmaW: autoregulation and partial overlap with the sigmaX regulon.
J Bacteriol 180: 3765–3770.26. Huang X, Gaballa A, Cao M, Helmann JD (1999) Identification of target
promoters for the Bacillus subtilis extracytoplasmic function sigma factor, sigmaW. Mol Microbiol 31: 361–371.
27. Hughes KT, Mathee K (1998) The anti-sigma factors. Annu Rev Microbiol 52:
231–286.28. Jervis AJ, Thackray PD, Houston CW, Horsburgh MJ, Moir A (2007) SigM-
responsive genes of Bacillus subtilis and their promoters. J Bacteriol 189: 4534–4538.
29. Jordan S, Hutchings MI, Mascher T (2008) Cell envelope stress response in
Gram-positive bacteria. FEMS Microbiol Rev 32: 107–146.30. Kanehara K, Ito K, Akiyama Y (2002) YaeL (EcfE) activates the sigma(E)
pathway of stress response through a site-2 cleavage of anti-sigma(E), RseA.Genes Dev 16: 2147–2155.
31. Kingston AW, Subramanian C, Rock CO, Helmann JD (2011) A sigma(W) -dependent stress response in Bacillus subtilis that reduces membrane fluidity.
Mol Microbiol 81: 69–79.
32. Leskela S, Kontinen VP, Sarvas M (1996) Molecular analysis of an operon inBacillus subtilis encoding a novel ABC transporter with a role in exoprotein
production, sporulation and competence. Microbiology 142: 71–77.
33. Luo Y, Helmann JD (2009) Extracytoplasmic function sigma factors with
overlapping promoter specificity regulate sublancin production in Bacillus
subtilis. J Bacteriol 191: 4951–4958.
34. Mascher T, Hachmann AB, Helmann JD (2007) Regulatory overlap and
functional redundancy among Bacillus subtilis extracytoplasmic function sigma
factors. J Bacteriol 189: 6919–6927.
35. Nicolas P, Mader U, Dervyn E, Rochat T, Leduc A, et al (2012) Condition-
dependent transcriptome reveals high-level regulatory architecture in Bacillus
subtilis. Science 335: 1103–1106.
36. Ogura M, Tsukahara K, Hayashi K, Tanaka T (2007) The Bacillus subtilis
NatK-NatR two-component system regulates expression of the natAB operon
encoding an ABC transporter for sodium ion extrusion. Microbiology 153: 667–
675.
37. Padan E, Schuldiner S (1994) Molecular physiology of Na+/H+ antiporters, key
transporters in circulation of Na+ and H+ in cells. Biochim Biophys Acta 1185:
129–151.
38. Petersohn A, Brigulla M, Haas S, Hoheisel JD, Volker U, et al (2001) Global
analysis of the general stress response of Bacillus subtilis. J Bacteriol 183: 5617–
5631.
39. Pietiainen M, Gardemeister M, Mecklin M, Leskela S, Sarvas M, et al (2005)
Cationic antimicrobial peptides elicit a complex stress response in Bacillus
subtilis that involves ECF-type sigma factors and two-component signal
transduction systems. Microbiology 151: 1577–1592.
40. Rasmussen S, Nielsen HB, Jarmer H (2009) The transcriptionally active regions
in the genome of Bacillus subtilis. Mol Microbiol 73: 1043–1057.
41. Saito A, Hizukuri Y, Matsuo E, Chiba S, Mori H, et al (2011) Post-liberation
cleavage of signal peptides is catalyzed by the site-2 protease (S2P) in bacteria.
Proc Natl Acad Sci U S A 108: 13740–13745.
42. Schobel S, Zellmeier S, Schumann W, Wiegert T (2004) The Bacillus subtilis
sigmaW anti-sigma factor RsiW is degraded by intramembrane proteolysis
through YluC. Mol Microbiol 52: 1091–1105.
43. Steil L, Hoffmann T, Budde I, Volker U, Bremer E (2003) Genome-wide
transcriptional profiling analysis of adaptation of Bacillus subtilis to high salinity,
J Bacteriol 185: 6358–6370.
44. Sterberg S, del Peso-Santos T, Shingler V (2011) Regulation of alternative sigma
factor use. Annu Rev Microbiol 65: 37–55.
45. Tojo S, Matsunaga M, Matsumoto T, Kang CM, Yamaguchi H, et al (2003)
Organization and expression of the Bacillus subtilis sigY operon. J Biochem 134:
935–946.
46. Turner MS, Helmann JD (2000) Mutations in multidrug efflux homologs, sugar
isomerases, and antimicrobial biosynthesis genes differentially elevate activity of
the sigma(X) and sigma(W) factors in Bacillus subtilis. J Bacteriol 182: 5202–
5210.
47. Wadenpohl I, Bramkamp M (2010) DivIC stabilizes FtsL against RasP cleavage.
J Bacteriol 192: 5260–5263.
48. Wiegert T, Homuth G, Versteeg S, Schumann W (2001) Alkaline shock induces
the Bacillus subtilis sigma(W) regulon. Mol Microbiol 41: 59–71.
49. Yoshimura M, Asai K, Sadaie Y, Yoshikawa H (2004) Interaction of Bacillus
subtilis extracytoplasmic function (ECF) sigma factors with the N-terminal
regions of their potential anti-sigma factors. Microbiology 150: 591–599.
50. Zweers JC, Wiegert T, van Dijl JM (2009) Stress-responsive systems set specific
limits to the overproduction of membrane proteins in Bacillus subtilis. Appl
Environ Microbiol 75: 7356–7364.
The Bacillus SigW Regulon
PLOS ONE | www.plosone.org 14 November 2012 | Volume 7 | Issue 11 | e48471
Top Related